A model of the diffusion of methane into a wellbore in overbalance with Oil Based Mud is described. In some situations, methane diffusion may cause unexpected gas in riser and well control problems. The situation when the OBM contaminated with diffused methane is circulated is simulated. The model has been used to analyze measurements of hydrocarbon obtained during drilling and circulation in an off-shore well in the North Sea, showing promising agreement with the model.
The tools used are: a) a finite difference diffusion program that simulates the diffusion process of methane molecules from the formation, through the mud invasion zone and into the base oil in the well bore, and b) a finite difference dynamic hydraulics model with modules dealing with the interaction of hydrocarbons entering the mud from the formation and the base oil in the mud.
The lighter hydrocarbons in OBM from the well were measured using FID gas chromatographs capable of accurately measuring the amount of C1 through C5 continuously. The measurements were analyzed using a PVT simulation program.
The diffusion was simulated in two positions in the well; the two days waiting for the 13 5/8" liner, 1850 m TVD; and the four days waiting for pipe at full depth, 4500 m TVD. The inputs to the diffusion model were based on measurements obtained while drilling in the two depths. The hydrocarbon measurements from circulations after the two and four days were compared with the amounts provided by the diffusion theory.
The distribution of the light hydrocarbon elements obtained during drilling through the formation and circulation after the delay periods, were compared. After the delay period, the relative amount of methane (compared with ethane, propane, etc.) was substantially greater. This is because methane has greater diffusion coefficient than the "larger" hydrocarbon elements.
The potential problems of having diffused gas in the base oil are shown using a hypothetical well where the OBM has been left overbalanced for days in a one kilometer long horizontal reservoir section. The contaminated OBM is circulated through the riser, and the low level of methane produces a domino-effect as it starts to boil out in the riser.
In-line gas-chromatograph measurements have provided information that shows the diffusion of methane into OBM while overbalanced.
This model used together with in-line gas-chromatograph measurements will provide additional insight in how light hydrocarbons end up in Oil Based Mud. The dynamic hydraulics model can be used to show the safety limit when leaving the well without circulation.
Connector joint strength and leak resistance depend on internal and external pressures. Axial tension or compression opens (dilates) the connector, while the pressures close the connector. This paper presents a "toy connector" model that incorporates actual connector elements, but in a simplified form useful for analysis, so equations of joint strength and leak resistance can be determined as functions of axial load and internal/external pressures. The model and example case in this paper confirm that dilatancy occurs and hydrostatic pressure plays a critical role in connector failure (leak, pull-out, or thread shear). Von Mises stress alone cannot model dilatancy and the hydrostatic effect.
Phillips, Anthony (Baker Hughes a GE Company) | Gavia, David (Baker Hughes a GE Company) | Noel, Afton (Baker Hughes a GE Company) | Savage, Michael (Baker Hughes a GE Company) | Stefanik, Thomas (Baker Hughes a GE Company)
In-bit vibration sensors reveal that stick-slip and lateral dysfunction are the primary cause of poor polycrystalline diamond compact (PDC) bit performance in the 12 ¼ -in section of the Midland basin. Premature dulling as a result of vibrations leads to low penetration rates and failure to reach the targeted depth. New drill bit technologies have been designed to mitigate stick slip and lateral vibrations. This paper shows the method utilized to reduce drilling vibrations and increase penetration rate and footage to meet the operator's objectives.
Performance benchmarks were established by conducting a post-run analysis of drilling parameters incorporating in-bit sensing vibration data and through PDC bit dull evaluation from offset runs. This analysis led to new designs incorporating shaped diamond elements (SDE) to mitigate lateral vibrations and ovoid's with adaptive exposure that mitigate stick-slip without hindering rate of penetration (ROP). After the field runs showed a poor dull on the 6-blade frame and low performance from the 7-blade frame, they were tested in multiple states in a high-pressure downhole simulator to determine which elements have the highest effect on performance. The learnings from these where applied to a more durable, higher performing design.
The 6-bladed bit drilled 6636 feet, which is average footage, compared to offsets. However, the vibrations recorded during the 6-bladed bit run were significantly lower than the typical offsets, 82% of the drilling hours registering with low levels compared to the 52% smooth drilling typically seen in this interval. The bit was tripped short of target depth, due to low ROP with a ring-out in the shoulder. To improve durability the team recommended a 7-blade PDC bit, which resulted in low performance through ROP. With this result, the bit was laboratory tested replicating the ROP observed at the end of the field run by selecting an equivalent carbonate rock and adjusting the simulator to the overburden pressure. This provided a baseline that could evaluate the true impact of design changes, on field performance. Comparison of the all the features indicate that edge geometry, blade count, cutter size and backrake angle can increase ROP by making the bit drill more efficient while decreasing overall bit aggressiveness at lower ROP's. The findings show the primary benefit is improved ROP at high power levels, and reduced bit reactive torque at low power levels, at the time when lateral vibrations typically occur. The adaptive and SDE features will further add to this performance by reducing vibrations. This holistic approach allowed the team to identify the primary performance limiters through field, laboratory and downhole vibration analysis. The suite of full bit simulator tests established several key learnings, to improve performance. These learning when applied to the new design, proved to have good durability, reduced vibration and high performance, meeting the customer's objectives.
Design improvements were achieved based on the results of field tests and a series of full bit high-pressure simulator tests. The combination of adaptive PDC drill bit technology and shaped diamond elements was used to reduce downhole vibrations, thereby enabling the operator to improve overall bit performance and durability.
The BP project team has considered increased reserves recovery by lowering the reservoir abandonment pressure below the initial design value. Through a multi-disciplinary approach, design assumptions and equipment ratings were systematically reviewed to determine which aspects factored into the decision to change reservoir management. Collapse loading of the 10 in. production liner was identified as a key variable.
The conventional design factor, a ratio of the design load to the API collapse rating, was deemed to be an insufficient way of characterizing design margin, primarily due to the perception of conservatism in the rating. While design factors are convenient for screening a casing string against an agreed-upon set of inputs and assumptions, there is little insight gained from comparing a 1.03 design factor to a 1.02 other than one value is higher than the other. The team embarked on a scope of work to characterize the probability of collapse as a function of reservoir abandonment pressure using reliability based design (RBD).
Physical testing was conducted to characterize the distribution of collapse resistance and the distribution of dimensional and strength parameters which govern collapse. The quality data sets are combined using the Klever-Tamano limit state equation to indirectly derive a distribution of collapse resistance. The destructive collapse tests provide both a direct measure of the distribution of collapse and a way to calibrate the limit state equation model uncertainty. Both the direct and indirect methods are useful in determining the probability of collapse for a design load.
Load uncertainty was characterized by considering variability of conditions across the wellstock, including depth, temperature and completion configuration. Casing wear was also considered in the assessment.
This paper outlines the RBD methodology used to support the decision to lower reservoir abandonment pressures. Details on how to construct the statistical collapse model are provided along with a discussion on interpretation and continuous improvement activities.
Wellbore intercept probability calculations are important not only for avoiding collisions but also for relief-well planning and planning for wellbore reentry in abandonment projects. This study reviews the current methods for calculating probability and their sensitivity to the input variables. A probability of intercept calculation is presented that is proven, robust, and can be implemented in a spreadsheet as input-wellbore positions and survey-error dimensions.
Given the lack of availability of empirical wellbore collision-frequency data, it was necessary to model the wellbore-intercept scenario using a Monte Carlo simulator constructed by breaking the two wellbores into short segments. The validity of the simulator is highly sensitive to the probability distribution of the input-survey errors. The input-probability distribution of well separations in the high side and lateral directions is determined from survey-comparison studies for both systematic and random-error sources. The simulator is used to compare the accuracy and reliability of different probability integration types and sensitivity to the input data and distribution of errors.
It is observed from survey comparisons that the distribution of survey inclination and azimuth errors behave as "heavy tailed" probability distributions such as Laplace or Student's t. This increases the probability of intercept at higher sigma levels than the previously assumed normal distribution.
The results of this study exhibit the sensitivity of the probability result to the degree of sophistication of the probability calculation from simple one-dimensional (1D) projection to three-dimensional (3D) integration of probability density. There is a law of diminishing returns with the sensitivity of complex models being lost in the reliability of the input-error values and assumptions on probability distributions.
One observation is that there is no "pedal curve" effect for high-angle well crossings. The simple point-to-point collision calculations using this vector method do not include the confidence in the relative direction of the two wellbores. Therefore, probability calculations and separation-factor calculations that engage the pedal-curve distances are overly conservative.
A reliable prediction of wellbore-intercept probability is achieved using a relatively simple method. This study shows the importance of the angle of intercept to enhance interception probability for relief wells and abandonment re-entry operations.
Weight on bit (WOB) is the axial trust force applied to the bit during drilling. It is one of the most important controllable variables of the drilling process. Still, it is quite rare and very difficult to measure it directly, and we are often left with estimating it indirectly through measurements of the tension at the top of the string, or even at the dead line anchor. This indirect WOB measurement, commonly called surface WOB, is most often calculated as measured off-bottom hook loadminus the actual hookload. This method assumes that the reference string weight is constant and does not change during drilling.
This paper challenges this assumption by discussing a variety of physical effects that can cause the reference hook load to change, during the drilling of a new stand and after a traditional WOB zeroing procedure is carried out. The paper discusses the following effects that can be used for correcting the reference hook load: 1)well bore friction, 2)added weight in air, 3)well inclination, 4)flow induced lift, 5) nozzle jet lift, 6)cuttings in suspension, and 7) back-pressure in managed pressure drilling operations.
Hohl, Andreas (Baker Hughes, a GE Company) | Herbig, Christian (Baker Hughes, a GE Company) | Arevalo, Pedro (Baker Hughes, a GE Company) | Reckmann, Hanno (Baker Hughes, a GE Company) | Macpherson, John (Baker Hughes, a GE Company)
Downhole tools in bottom-hole assemblies are subject to high dynamic loads during drilling operations. The negative impacts of these dynamic loads can be inefficient drilling with low rate of penetration (ROP), reduced downhole directional and formation measurement service quality, and downhole tool failures with associated non-productive time.
The dynamic phenomena can be categorized by direction into axial, torsional and lateral vibrations, and by excitation mechanism into forced excitation, self-excitation, and parameter excitation. Forced vibrations are mainly caused by imbalances in the drilling system or by the working principles of downhole tools such as the mud motor. Self-excitation mechanisms are mostly driven by the interaction of the bit, reamer or drilling system with the formation, and can cause detrimental dynamic behavior such as torsional stick-slip, bit bounce, and backward whirl.
These diverse vibration phenomena require tailored mitigation strategies. To a certain extent, these mitigation strategies are contradictory. Misinterpretation of downhole measurements can lead to even worse vibration levels with severe consequences for reliability, ROP, and measurement quality.
As a consequence, downhole measurement devices should differentiate vibration phenomena. This distinct differentiation could then be used to choose appropriate mitigation strategies.
This paper analyses and defines the requirements for dynamics measurement devices. The specification, number, and placement of sensors and their associated sampling rates are examined to distinguish vibration directions and phenomena. The usefulness of these requirements is demonstrated using examples of torsional stick-slip and high-frequency torsional oscillations, lateral vibrations, and backward whirl. The results of kinematic modeling are analyzed and compared to high-speed vibration data from field runs measured with the latest generation of vibration measurement tools. Possible misinterpretation of vibration conditions in the case of inappropriate measurements is shown. The results are discussed by comparing theoretical modeling with field data.
The defined requirements and guidelines enable a flawless interpretation of downhole vibration measurements and unveil drilling optimization opportunities. Different vibration phenomena can be identified reliably and appropriate mitigation strategies applied in real time at the wellsite by the driller or automation systems. This finally reduces the vibration load on the drilling system, increasing its reliability and performance.
This paper serves to provide a technical overview of the Real-Time Drilling (RTD) analytics system currently developed and deployed. It also serves to share practices used in managing the RTD analytics project which have resulted in the efficient delivery of work products. By employing the novel and agile development approach on the RTD project, the design to production time has been faster and has cost much less compared to a more traditional multi-year effort and cost intensive RTD development project. Within three months, for proof of concept (PoC) purpose, an RTD analytics system with two analytics modules was built from scratch and placed online in production. This real-time decision-support tool has been fully accepted by the operations team and has become a powerful tool for daily well operations. After eleven months as this paper was drafted, this system has four analytic modules online for production and three analytic modules under development; it is expected that more new modules will be added to the system on a regular basis.
Lost circulation while running casing, pre-cement job mud circulation, and cement placement can result in additional remediation, create challenges in achieving zonal isolation objectives, and potentially impact well integrity. Several mitigation methods are available to address lost circulation challenges. However, the focus of this article is on prevention.
Wellbore strengthening is utilized to strengthen low fracture gradient (FG) and permeable zones to a target maximum equivalent mud weight (EMW) value induced during tripping, circulating, or cementing. Engineered wellbore strengthening pills of specific particle concentration and size distribution are placed in the open hole and pressure is applied to the wellbore in a controlled fashion. As a result, a higher fracture gradient of the weak sands exposed in the entire interval is achieved, allowing permeable formations to withstand higher EMW during subsequent tripping, circulating, and cementing operations without inducing losses. This paper summarizes the field application of wellbore strengthening in deep-water Gulf of Mexico (GoM) to achieve zonal isolation. The design, execution, and post-well analysis of several wellbore strengthening applications in a variety of sands from shallow and soft overburden sands to deep producing reservoir sands with depletions greater than 4,000 psi, are discussed. A detailed case study for one of the wells is provided.
The track record of successful wellbore strengthening treatments performed demonstrates the potential for achieving adequate annular cement placement and zonal isolation in narrow pore pressure fracture gradient environments.
Wellbore stability is one determining factor to the success of drilling operations. A multitude of wellbore stability issues may still be encountered, despite all previous knowledge and experience, when drilling a certain formation. However, the use of the large amount of data collected during drilling can provide valuable insight. This collected data, in combination with artificial intelligence models, can be used to predict wellbore stability in new drilling operations. Because wellbore stability is such a common and expensive issue to the industry, many papers in the literature have presented a multitude of techniques and algorithms to predict wellbore stability. The predictions made in the literature have all provided relatively good accuracy; the most common methods that have been presented include the use of multivariate statistics, linear methods, and artificial neural networks.
This paper presents two approaches to solve the wellbore instability prediction problem. First, principal component analysis (PCA) and linear discriminant analysis (LDA) are used in an attempt to reduce the dimensionality of the dataset. The results indicate that the dataset dimensionality could not be reduced. Second, nonlinear artificial intelligence models were implemented, including the artificial neural network (ANN), specifically the Bayesian regularization neural network (BRNN), and support vector machine (SVM). After building these two models and checking their accuracy, a new feature was implemented to filter the input data. Empirical mode decomposition (EMD) was adopted, which is commonly used in the field of signal processing to eliminate noise and variation in the data. This study evaluates the effect of EMD on the prediction and classification accuracy for wellbore stability. The study results are promising; by using EMD, accuracy was increased by 15% for both the ANN and the SVM models.